1. Field of the Invention
The present invention relates to a power supply apparatus and a printing apparatus. Particularly, the present invention relates to a power supply apparatus and a printing apparatus which prints an image on a print medium using an inkjet printhead, and using the power supply apparatus.
2. Description of the Related Art
Inkjet printers as computer peripheral devices have developed to multi-function printers including scanner and copy functions, and also a FAX function in recent years, and have higher degrees of penetration into the market year by year due to their higher cost performance ratios.
In general, a switching power supply is used as a power supply of an inkjet printer. A conventional dropper type power supply, which operates at a commercial frequency (50/60 Hz), is falling out of use since it is inferior in terms of power conversion efficiency, heat generation, and the like.
Switching power supplies are classified into some types based on their switching methods, and a flyback method is most popularly used in terms of its simple circuit arrangement and cost. Recently, many control ICs exclusively used for such flyback method have been provided from respective manufacturers, and a circuit having high reliability can be designed relatively easily.
FIG. 8 is a circuit diagram showing an example of a conventional switching power supply using the flyback method. In this example, an IC 1 is a commercially available PWM (Pulse Width Modulation) control IC, and implements feedback control of a current mode. An outline of the operation of the circuit shown in FIG. 8 will be described below.
An input voltage of a commercial frequency of 50 Hz or 60 Hz is rectified by a bridge diode BD1, and is then smoothed by an electrolytic capacitor C1 to generate a DC voltage Vin(DC). This DC voltage Vin(DC) is about 140 V in Japan, or is about 320 V in a 230-V region in Europe and the like. This DC voltage Vin(DC) is supplied to a transformer T1, and undergoes switching control by a transistor Q1. As a result, an energy accumulated on a primary side winding 11 of the transformer T1 is transferred to a secondary side winding 12, thereby generating a DC output voltage Vo.
More specifically, in the flyback method, an energy is accumulated on the primary side winding 11 (n1: the number of windings) of the transformer during a turn-on period of the transistor Q1 in FIG. 8, and the accumulated energy is transferred to the secondary side winding 12 (n2: the number of windings) during an off period of the transistor Q1. The energy transferred to the secondary side winding 12 in this way is rectified and smoothed by a diode D2 and electrolyte capacitor C4 to generate the DC output voltage Vo. The output voltage Vo is voltage-divided by resistors R6 and R7, and a divided voltage (a voltage of a node Vref1) is input to a reference terminal (REF) of a constant voltage regulator, that is, a so-called shunt regulator IC 3. FIGS. 15A and 15B are explanatory circuit diagrams of the shunt regulator IC 3. The shunt regulator IC 3 is one kind of so-called error amplifiers. In the example of FIG. 8, a cathode (K) is connected to a photocoupler IC 2, and an anode (A) is connected to ground. Therefore, an anode voltage Va is 0 V. The shunt regulator IC 3 has an error amplifier (comparator) 151 and a fixed reference voltage circuit 152 arranged in the regulator. The shunt regulator IC 3 compares a voltage of the node Vref1 input to the reference terminal (REF) with an output voltage of the fixed reference voltage circuit 152, and outputs a voltage Vk. In this example, the fixed reference voltage circuit generates 2.5 V. The cathode voltage Vk as an output voltage of the regulator IC 3 is controlled, so that the reference voltage Vref1 input to the reference terminal (REF) of the IC 3 always becomes DC 2.5 V, thus achieving feedback control. Please refer to specialized books and the like for the detailed operation of the shunt regulator since it is well known in this art, and a description thereof will not be given. Note that symbols R1, R2, R3 and R5 in FIG. 8 are resisters, C2 is another electrolytic capacitor.
For example, when the output voltage Vo rises, and the input voltage Vref1 of the shunt regulator IC 3 consequently rises, the output Vk of the shunt regulator IC 3 drops conversely. As a result, a current which flows through a resistor R9 and an LED 15 of the photocoupler IC 2 increases. Then, a collector current which flows through a phototransistor 16 of the photocoupler IC 2 increases, and a potential of a feedback terminal FB of the control IC 1 drops. Then, finally, a pulse width, in other words, an ON duty of a PWM signal output from a DRV terminal of the control IC 1 drops, and a turn-on time of the transistor Q1 is consequently shortened (conversely, a turn-off time is prolonged). As a result, an energy accumulated on the primary side winding 11 of the transformer decreases, that to be transferred to the secondary side winding 12 decreases accordingly, and the output voltage Vo drops finally.
In this way, when the output voltage Vo rises, the feedback control effects to cancel it. Conversely, when the output voltage Vo drops, the feedback control effects to raise that voltage, thereby obtaining the stable DC output voltage Vo. More specifically, a circuit bounded by a broken line 14 in FIG. 8 is a control circuit which is responsible for gain adjustment and phase adjustment required for such feedback control, and functions to stably operate the entire system. The control circuit 14 has an input node 14in and output node 14out. More specifically, a resistor R8 and capacitor C5 in the broken line 14 serve as such gain and phase adjustment parameters. Note that hereinafter, the circuit portion bounded by the broken line 14 will be referred to as an error amplifier.
The description will be further continued with reference to FIG. 8. The transformer T1 includes an auxiliary winding 13, which is used to generate a power supply voltage Vcc for the control IC 1. More specifically, a voltage generated by the auxiliary winding 13 is rectified and smoothed by a diode D1 and electrolyte capacitor C3, and is further stepped down by a transistor Q2 and Zener diode ZD1, thus generating the power supply voltage Vcc of the control IC 1. In the example of FIG. 8, Vcc=15 V, and hence, the Zener diode ZD1 also has a 15-V specification. Note that the control IC 1 is connected to the transistor Q1 via a resistor R4, and has a CS terminal required to detect a current flowing through the transistor Q1 and an HV terminal required to detect a voltage Vin (DC).
FIG. 9 is a block diagram showing an outline of the feedback control of the switching power supply shown in FIG. 8.
As shown in FIG. 9, a PWM control unit 80 including a dedicated IC and the like normally supplies a PWM control signal 81 to a driver 82 including a switching element (Q1), and the driver 82 drives a transformer 83. As a result, an energy is transferred to the output side of the transformer 83, and an output voltage Vo1 is generated via a rectifying and smoothing circuit 84 in the example shown in FIG. 8.
A framework of the feedback control will be described below with reference to FIG. 8 again. A voltage variation of the output voltage Vo1 is detected as a feedback current If1(dc). In the circuit shown in FIG. 8, the current If1(dc) is given by:If1(dc)=(Vo1−Vref)/R6  (1)where Vref is a reference voltage of the reference terminal (REF) of the shunt regulator IC 3, and is, for example, DC 2.5 V. This feedback current If1(dc) flows into the reference node Vref1. On the other hand, a current Iref1 which flows out from the node Vref1 is given by:Iref1=Vref/R7  (2)
Then, since the entire system is controlled so that If1(dc) and Iref1 become equal to each other, equation (3) is obtained from equations (1) and (2). That is, we have:Vo1=(R6+R7)/R7*Vref  (3)In this way, the output voltage Vo1 is controlled based on equation (3).
Referring back to FIG. 9, the description will be continued. A feedback factor α1 corresponds to a coefficient when the output Vo1 is considered as a variable in equation (1) which defines the feedback current If1(dc), and is 1/R6. A degree D(α1) at which the feedback factor α1 contributes to the feedback control is 1.0. This is because the circuit shown in FIG. 8 includes only one output as a feedback control target. By contrast, as will be described later, in case of a power supply having two outputs, since contributions of feedback factors are weighted between the two outputs, degrees D(αn) assume values which satisfy 0<D(αn)<1 (n=1, 2). Referring back to FIG. 9, the feedback signal weighted by a weighting circuit 86 is input to an error amplifier 89, and its output is provided to the PWM control unit 80, thus executing the aforementioned PWM control.
Details of the operation of the switching power supply shown in FIG. 8 will be described below with reference to waveforms of the respective units.
FIG. 10 is a signal waveform chart showing a drain-source voltage Vds and drain current Id of the transistor Q1, a current which flows through the secondary side winding 12 of the transformer, that is, a current Is which flows through the rectifier diode D2, and an output current Io in the switching power supply.
Note that FIG. 10 shows, as a representative example, waveforms of a current-discontinuous mode of the flyback method. Note that as is apparent to an ordinary skilled person in the art, when a load power increases, and a PWM ON duty becomes equal to or higher than 50%, the current-discontinuous mode set so far transits to a current-continuous mode. However, such mode transition is not directly related to the gist of the present invention, and a description thereof will not be given.
In FIG. 10, a basic cycle T of a switching operation is, for example, 16.7 μsec when the operation frequency is 60 kHz. This cycle includes a period Ton in which the transistor Q1 is turned on, and a period Toff in which the transistor Q1 is turned off. Furthermore, the period Toff includes a period Toff1 in which an energy is discharged from the secondary side winding 12 of the transformer via the diode D2 and electrolyte capacitor C4, and a standby period Toff2 after discharging is complete and until the transistor Q1 is turned on again. In the period Toff2, as can be seen from FIG. 10, the drain-source voltage of the transistor Q1 resonates. This is a phenomenon generally caused by a resonance system formed by an inductance value L1 of the primary side winding 11 of the transformer, a leakage inductance value Lleak, and a total capacitance value Clump between the drain and source of the transistor Q1. However, since this phenomenon is not directly related to the gist of the present invention, a detailed description thereof will not be given.
In the current-discontinuous mode shown in FIG. 10, an energy is accumulated on the primary side winding 11 of the transformer during the period Ton. This energy is given by:P1=1/2*L1*Ip2  (4)where L1 is an inductance value of the primary side winding 11, and Ip is a peak value of a current which flows through the primary side winding 11 during the period Ton, as shown in FIG. 10.
Next, an energy generated by the transformer per unit time is described by:P2=1/2*L1*Ip2*f*η  (5)where f is a switching frequency, and η indicates energy conversion efficiency of the transformer. A product of the energy amount P1 generated on the primary side of the transformer and the efficiency η is an energy amount which is actually transferred to the secondary side of the transformer. For example, f is 60 kHz, 100 kHz, or the like, and η is 0.95 or the like.
That is, about 95% of the energy generated on the primary side of the transformer is transferred to the secondary side, and the remaining 5% is dissipated as heat by a core and the windings of the transformer. Note that for reference, the total efficiency of a switching power supply of the flyback method of several ten W output is about 85%. In addition to the aforementioned dissipation by the transformer alone, that by an EMI filter circuit (not shown) of an input unit, that by the switching element Q1, that by the rectifier D2 in the secondary side circuit, that by resistors in the circuit shown in FIG. 8, and the like are included.
The description will be continued with reference to FIG. 10 again. In equation (4), the energy P1 accumulated during the period Ton is transferred to the secondary side winding 12 of the transformer during the period Toff1. This switching control method will be referred to as a flyback method hereinafter. By contrast, a method in which an energy is transferred from the primary side to the secondary side of the transformer during the period Ton of the transistor Q1 is available, and will be referred to as a forward method hereinafter. Please refer to specialized books for details of the forward method.
In the circuit shown in FIG. 8, one output Vo (for example, DC 24 V) is generated as an output voltage. However, in an inkjet printer, as shown in, for example, FIG. 11, the output voltage Vo is supplied to a printhead 3 and motor driver 44, and is also supplied to a DC-DC converter 45 which is used to generate several types of logic circuit voltages.
FIG. 11 is a block diagram showing the arrangement of a power supply unit of the inkjet printer.
These logic circuit voltages include, for example, DC 1.5 V used as voltages of a CPU and ASIC core, DC 3.3 V supplied to an ASIC input/output unit (I/O) and memory device, DC 5V supplied to sensors and a display unit, and the like. Note that as shown in FIG. 11, motors connected to the motor driver 44 include a conveyance motor M2, and a carriage motor M1 which drives a carriage which mounts the printhead 3 and is scanned. Furthermore, in a recent multi-function printer (MFP), the motors include a scanner (SC) motor M3 used to scan a scanner unit and the like. In order to meet a recent power-saving requirement, an energy-saving control signal (Esave) required to guide a power supply circuit 42 to intermittent oscillations in a standby state or sleep state of a printer is normally included.
In the arrangement example shown in FIG. 11, both of a driving voltage of the printhead 3 and a voltage of the motor driver 44 are indicated by one output voltage DC 24 V. However, due to recent high-speed trends of printers, a case in which higher voltages DC 27 V, DC 32V, and the like are used as a motor driving voltage is increasing. In such case, two outputs, that is, a printhead driving voltage (DC 24 V) and motor driving voltage (for example, DC 32 V) are generated.
FIG. 12 is a circuit diagram showing an example of a switching power supply having two output voltages. The same symbols and the same reference numerals in FIG. 12 denote components common to FIG. 8.
A major difference from the circuit shown in FIG. 8 lies in that two types of output voltages Vo1 and Vo2 are generated. For example, the output Vo1 is DC 24 V corresponding to a head driving voltage, and the output Vo2 is DC 32 V corresponding to a motor driving voltage. In order to generate these two types of output voltages, two windings 12 (n2: the number of windings) and 12a (n3: the number of windings) are provided to the secondary side of a transformer T9, so that the output voltage Vo1 is generated from the winding 12 (n2: the number of windings), and the output voltage Vo2 is generated from the winding 12a (n3: the number of windings). Note that symbols Is1, Is2 represent the output current from the windings 12, 12a, respectively. As is apparent to an ordinary skilled person in this art, the windings 12 and 12a may be configured via an intermediate tap 51 (the windings 12 and 12a share a specific pin terminal of the transformer), or the respective windings may be independently wound.
As a rectifying and smoothing circuit from the added winding 12a, a diode D51 and electrolyte capacitor C51 are arranged. To the input terminal (REF) of the shunt regulator IC 3, a feedback from the output voltage Vo2 is input in addition to that from the output voltage Vo1. In this case, the feedback from the output voltage Vo1 is attained by DC coupling via the resistor R6, but that from the output voltage Vo2 is attained by AC coupling by means of a resistor R51 and capacitor C52 (a broken line 52 in FIG. 12).
This is because since the output voltage Vo1 is used as a driving voltage of the printhead, as described above, very high precision control is required, while since the output voltage Vo2 is a voltage used to drive a DC motor and the like, variations to some extent are tolerated. Therefore, as for the output voltage Vo2, the feedback by means of the AC coupling is used so as to avoid an extreme voltage drop at a timing at which a large current is instantaneously supplied (for example, a motor activation timing). In other words, in the example of the circuit shown in FIG. 12, the feedback of the output voltage Vo1 is always prioritized. On the other hand, as for the output voltage Vo2, feedback control according to a feedback factor decided by a CR time constant of the resistor R51 and capacitor C52 is implemented only for instantaneously large load variations.
Note that Japanese Patent Laid-Open No. 6-178537 is available as an example of a related art associated with feedback control of a switching power supply which generates two output voltages.
Japanese Patent Laid-Open No. 6-178537 discloses a method of selecting an output voltage used as a feedback control target in accordance with respective load currents of a plurality of output voltages. According to Japanese Patent Laid-Open No. 6-178537, an output voltage with a larger load current is selected to execute feedback control.
However, the conventional circuit shown in FIG. 12 suffers the following problems. That is, since the feedback of the output voltage Vo2 is attained by the AC coupling, even though a printhead driving operation is normally set in a quiescent period during, for example, a high-speed print medium conveyance execution period in a print operation sequence, the feedback of the printhead voltage Vo1 is unwantedly prioritized. For this reason, the motor driving voltage Vo2 largely varies consequently, thus adversely influencing motor servo-control. Thus, in practice, high-speed control of motors is implemented within a range in which stability of such servo-control is maintained in the conventional circuit.
On the other hand, in the circuit shown in FIG. 12, the feedback of the output Vo2 corresponding to the motor driving voltage may also be attained by DC coupling as in the output Vo1. In this case, a relative feedback factor ratio between the respective outputs has to be decided, and when a large ratio is set for the output Vo1 corresponding to the printhead voltage, the stability of the output Vo1 is maintained. However, at a motor driving or stop timing, variations caused by overshooting or undershooting of the motor driving voltage Vo2 become larger than those in the case of the AC coupling. Conversely, a large feedback factor is set for the output Vo2, variations of the printhead voltage Vo1 increase, thus seriously deteriorating image quality.
FIG. 13 is a signal waveform chart showing voltage and current waveforms of the respective units in the 2-output voltage switching power supply shown in FIG. 12.
For example, at an activation timing (t=t1) of the conveyance motor M2, when a load current Io2 of the output Vo2 corresponding to the motor driving voltage is reduced by a peak current Ip1, the output Vo2 drops (undershoots) to Vp21 at that instance. A feedback current If2(ac) shown in FIG. 13 flows out from the node Vref1 toward the CR circuit 52, and the potential of the node Vref1 drops by that current flow-out amount.
As a result, the output voltage Vk of the shunt regulator IC 3 rises, and the ON duty of the PWM control by the control IC 1 increases to increase an energy generated by the transformer T9, thus effecting the feedback control that blocks the output voltage Vo2 from dropping. However, in the circuit shown in FIG. 12, even after the output current Io2 is stabilized to Ip2 at time t=t2, the output voltage Vo2 unwantedly drops by ΔVf compared to a level Vp20 before time t=t1, thus becoming a level Vp22.
This is because, as can be seen from the arrangement shown in FIG. 12, during a stable period of the output voltage Vo2, the feedback of that output does not function due to the AC coupling, and only the feedback of the output Vo1 is virtually effective. In other words, since the feedback of the output Vo2 is uncontrolled in terms of DC during a period T2, it exhibits a voltage drop tendency as its load current increases. For example, during a high-speed print medium conveyance period by the conveyance motor M2 shown in FIG. 11, the voltage drop ΔVf of the motor driving voltage Vo2 reaches about 3 to 4 V.
The description will be continued with reference to FIG. 13 again. At time t=t3, the driving operation of the conveyance motor M2 is stopped, and the current Io2 returns to zero. The output Vo2 instantaneously rises (overshoots) to Vp23 accordingly, and the feedback current If2(ac) reversely flows in from the CR circuit 52 to the node Vref1. As a result, the potential of the node Vref1 rises, the output voltage Vk of the shunt regulator IC 3 drops, and the ON duty of the PWM control by the control IC 1 decreases, thus reducing an energy generated by the transformer T9. In this manner, the feedback control that blocks the output voltage Vo2 from rising is effected.
Note that as shown in FIG. 13, the other output Vo1 is influenced by the large variations of the voltage Vo2 at times t=t1 and t3, and changes slightly in directions opposite to the voltage Vo2. That is, at instances of the large variations of the output Vo2, the output current Io1 of the output Vo1 exhibits only moderate changes, as shown in FIG. 13. For this reason, as described above, due to transfer of a large energy from the transformer by the feedback control caused by variations of the output Vo2, the output voltage Vo1 rises slightly at time t=t1, and reaches a Vp11 level, as shown in FIG. 13. Likewise, at time t=t3, the output Vo1 slightly drops to a Vp12 level conversely.
FIG. 14 is a block diagram showing the feedback control arrangement of the 2-output voltage switching power supply shown in FIG. 12. A difference from the 1-output voltage switching power supply described above with reference to FIG. 9 lies in that the transformer 83 has two outputs (83a and 83b), which are generated as the outputs Vo1 and Vo2 via rectifying and smoothing circuits 84a and 84b, respectively, as shown in FIG. 14. Also, respective feedback components from the outputs Vo1 and Vo2 are added by an adder 88 via weighting circuits 86, 87 which respectively weight feedback components by feedback factors α1 and α2, and the addition result is processed by the error amplifier 89 and is finally fed back to the PWM control unit 80. Thus, the driver 82 including the switching element (Q1) undergoes PWM control, thus controlling an energy to be generated by the transformer 83.
Especially, a change of the output voltage Vo1 is reflected to the feedback control via the feedback factor α1, and a change of the output voltage Vo2 is reflected to the feedback control via the feedback factor α2. As will be described later, these feedback factors α1 and α2 correspond to degrees of contribution of the corresponding outputs to the feedback control.
The description will be continued with reference to FIGS. 12 and 14. Since the feedback of the output voltage Vo1 is attained by the DC coupling, the constant feedback current If1(dc) always flows in from an output terminal Vo1 toward a reference terminal Vref1 of the shunt regulator IC 3. The current If1(dc) is given by:If1(dc)=(Vo1−Vref)/R6  (6)where the reference voltage Vref is, for example, DC 2.5 V. On the other hand, since the feedback of the output voltage Vo2 is attained by the AC coupling, the feedback current If2(ac) which flows in from an output terminal Vo2 to the node Vref1 is cut-off to zero by the capacitor C52 during a stable period of the output voltage Vo2. Therefore, a total sum of currents which flow into a node Vref1 in a Vo2 stable period is given by:If1(dc)+If2(ac)≈If1(dc)(∵If2(ac)≈0)  (7)
Therefore, expression (8) is obtained from equation (6) and expression (7). That is, we have:If1(dc)+If2(ac)≈(Vo1−Vref)/R6  (8)
Also, from expression (8), the feedback factors α1 and α2 are calculated, as given by:α1=1/R6, α2=0  (9)This is because when the output voltages Vo1 and Vo2 are considered as variables in expression (8), their coefficients correspond to the feedback factors. However, in practice, the variable Vo2 does not appear in expression (8), and α2 becomes zero.
Next, from equations (9), feedback contributions ratios D(α1) and D(α2) of α1 and α2 are calculated, as given by:
                                                                        D                ⁡                                  (                                      α                    ⁢                                                                                  ⁢                    1                                    )                                            =                            ⁢                              α                ⁢                                                                  ⁢                                  1                  /                                      (                                                                  α                        ⁢                                                                                                  ⁢                        1                                            +                                              α                        ⁢                                                                                                  ⁢                        2                                                              )                                                                                                                          =                            ⁢              1.0                                                          (        10        )                                          D          ⁡                      (                          α              ⁢                                                          ⁢              2                        )                          =        0                            (        11        )            
As can be seen from the above description, only the output Vo1 contributes to the feedback during a stable period of the output voltage Vo2 (the period T2 in FIG. 13).
On the other hand, as shown in FIG. 12, the current Iref1 which flows out from the node Vref1 is given by:Iref1=Vref/R7  (12)Since the current which flows into the node Vref1 and that which flows out from that node are controlled to be equal to each other by the operation of the error amplifier 14 including the shunt regulator IC 3, equation (13) holds. That is, we have:If1(dc)+If2(ac)=Vref/R7  (13)In this case, expression (14) is obtained from expression (7). That is, we have:If1(dc)≈Vref/R7  (14)
On the other hand, when the motor driving voltage instantaneously varies like in a period T1 in FIG. 13 (for example, at activation timings of the conveyance motor M2 and carriage motor M1), the output Vo2 is reduced by a voltage which contributes to supplying a large current. For this reason, if the feedback of the output Vo2 is not taken into consideration, the output voltage Vo2 causes a large voltage drop at that instance, and the servo-control of motors may become abnormal. For this reason, since the circuit shown in FIG. 12 adopts the feedback by means of the AC coupling of the output Vo2, the following feedback correction is executed against such instantaneous variations of the output Vo2. That is, letting ΔVp be a voltage variation of the output Vo2, the feedback current If2(ac) from the output Vo2 flows out from the node Vref1 to the CR circuit 52 (see FIG. 12)If2(ac)=−ΔVp/R51*exp(−T/CR)  (15)
In this case, the minus sign in equation (15) means flowing-out of the current from the node Vref1. Also, C of the CR time constant in equation (15) is a value of the capacitor C52, R is that of the resistor R51, and a variable T corresponds to an elapsed time since the motor activation timing, that is, an elapsed time from time t=t1 in FIG. 13. Furthermore, the period T1 of Vref1 shown in FIG. 13 reveals a potential drop state of the node Vref1 by a flow-out amount of the current If2(ac). In FIG. 13, Ifp21 shows a minimum value of the current If2(ac) during a time interval T1, and Ifp 22 shows a maximum value of the current If2 (ac) during a time interval T3. Also, in FIG. 13, a time interval T1 from times t=t1 to t2 is associated with the CR time constant in equation (15).
As described above, during the period T1 in FIG. 13, from equations (6) and (15), a total sum of feedback currents is calculated, as given by:If1(dc)+If2(ac)=(Vo1−Vref)/R6−ΔVp/R51*exp(−T/CR)  (16)From equation (16), respective coefficients of the variables Vo1 and ΔVp, that is, feedback factors are respectively calculated, as given by:α1=1/R6  (17)α2=−1/R51*exp(−t/CR)  (18)
In equation (18), the minus sign of α2 indicates a drop of the output voltage Vo2, and conversely, a plus sign indicates a rise of the output voltage Vo2. Also, as can be seen from equation (18), the feedback factor at time t=0 is larger as the resistor R51 in FIG. 12 decreases.
That is, the current which flows out from the node Vref1 increases with decreasing resistor R51, and a change of the output voltage Vo2 is consequently reflected to the node Vref1 at a high sensitivity, thus transferring the change to the subsequent error amplifier 14. However, when the value of the resistor R51 is set to be too small, the influence of the feedback factor α2 becomes too large, thus causing excessive overshooting or undershooting at a return timing from variations of the output voltage Vo2, and adversely influencing the other output Vo1.
Also, as can be seen from equation (18), the feedback factor of the output Vo2 includes an exponential function. For this reason, for example, when there are a plurality of motors (carriage motor M1, conveyance motor M2, scanner motor M3, and the like) of different activation currents, it is difficult to optimally control all the motors using one type of the CR time constant.
As described above, in the conventional 2-output voltage switching power supply shown in FIG. 12, an optimal feedback factor cannot be selected at the time of instantaneous variations of the motor driving voltage caused at activation timings of the plurality of motors having different activation currents. Furthermore, the feedback factor α2 of the motor driving voltage becomes zero in effect after the motors reach nearly a stabilized condition, thus consequently causing a drop of that output voltage. As a result, in the conventional circuit shown in FIG. 12, the printhead driving voltage can be precisely maintained, while the motor driving voltage unwantedly drops in a high-speed print medium conveyance period or the like. Hence, it is difficult to attain a high-speed throughput.